Kouichi Doi | PubMed Health

PubMed2005

Gene/protein name recognition based on support vector machine using dictionary as features.

BACKGROUND: Automated information extraction from biomedical literature is important because a vast amount of biomedical literature has been published. Recognition of the biomedical named entities is the first step in information extraction. We developed an automated recognition system based on the SVM algorithm and evaluated it in Task 1.A of BioCreAtIvE, a competition for automated gene/protein name recognition. RESULTS: In the work presented here, our recognition system uses the feature set of the word, the part-of-speech (POS), the orthography, the prefix, the suffix, and the preceding class. We call these features "internal resource features", i.e., features that can be found in the training data. Additionally, we consider the features of matching against dictionaries to be external resource features. We investigated and evaluated the effect of these features as well as the effect of tuning the parameters of the SVM algorithm. We found that the dictionary matching features contributed slightly to the improvement in the performance of the f-score. We attribute this to the possibility that the dictionary matching features might overlap with other features in the current multiple feature setting. CONCLUSION: During SVM learning, each feature alone had a marginally positive effect on system performance. This supports the fact that the SVM algorithm is robust on the high dimensionality of the feature vector space and means that feature selection is not required.

Algorithms↗

PubMed2005

Viewing the proteome from oligopeptides and prediction of protein function.

Our research activity of making the lexicon of relatively short oligopeptides has been one of the first steps to view the world of proteome from the perspective of oligopeptides. We propose a new method for the prediction of protein function, especially GeneOntology terms (GO terms), based on statistical characteristics of oligopeptides as an application of the lexicon. In the lexicon, a known function of a protein is inherited to its oligopeptides, and the correspondence between oligopeptides and the function is calculated in the whole proteins. In our method, unknown functions of proteins are predicted by means of the correspondence automatically. We measured the prediction performance using the 28,520 whole human proteins registered in RefSeq for several GO terms by recall-precision graphs. The GO terms include 'membrane', 'nucleus', 'ATP binding', 'hydorolase activity', 'GTP binding', 'intracellular signaling cascade' and 'ubiquitin cycle'. In most cases, it scores 70% recall with 80% precision. The prediction for ATP binding and GTP binding results in quite high performance: it scores 80% recall with 80% precision. Even in the worst case (ubiquitin cycle), it scores 62.6% recall with 80% precision. These results suggest that the proposed method is quite efficient for predicting GO terms.

Computational Biology↗